Second-generation rabies virus-based vaccine vectors expressing human immunodeficiency virus type 1 gag have greatly reduced pathogenicity but are highly immunogenic

Abstract

Rabies virus (RV) vaccine strain-based vectors show great promise as vaccines against other viral diseases such as human immunodeficiency virus type 1 (HIV-1) infection and hepatitis C, but a low residual pathogenicity remains a concern for their use. Here we describe several highly attenuated second-generation RV-based vaccine vehicles expressing HIV-1 Gag. For this approach, we modified the previously described RV vaccine vector SPBN by replacing the arginine at position 333 (R333) within the RV glycoprotein (G) with glutamic acid (E333), deleting 43 amino acids of the RV G cytoplasmic domain (CD), or combining the R333 exchange and the CD deletion. In addition, we constructed a new RV vector that expresses HIV-1 Gag from an RV transcription unit upstream of the RV phosphoprotein gene (BNSP-Gag) instead of upstream of the G gene. As expected and as demonstrated for SPBN-Gag, all vaccine vehicles were apathogenic after peripheral administration. However, the new, second-generation vaccine vectors containing modifications in the RV G were also apathogenic after intracranial infection with 10(5) infectious particles, and BNSP-Gag produced a 50%-reduced mortality in mice. Of note, the observed attenuation of pathogenicity did not result in either the attenuation of the humoral response against the RV G or the previously observed robust cellular response against HIV-1 Gag. These findings demonstrate that very safe and highly effective RV-based vaccines can be constructed and further emphasize their potential utility as efficacious antiviral vaccines.

FIG. 1.

Construction of different recombinant vaccine vectors expressing HIV-1 Gag. At the top are two RV vaccine strain-based vectors containing an additional transcription stop-start signal flanked by two unique restriction sites between the G and L genes (SPBN) or the N and P genes (BNSP). These vectors were the targets to introduce the gene encoding HIV-1 Gag (SPBN-Gag, SPBN-333-Gag, SPBN-ΔCD-Gag, SPBN-333-ΔCD-Gag, and BNSP-Gag). By site-directed mutagenesis and a PCR strategy, an amino acid change at position 333 within the RV G from arginine (R) to glutamic acid (E) was introduced (SPBN-333-Gag and SPBN-333-ΔCD-Gag) and/or 43 of 44 aa of the RV G CD (black boxes) were deleted (SPBN-ΔCD-Gag and SPBN-333-ΔCD-Gag).

FIG. 2.

Western blot analysis of recombinant vaccine vectors. BSR cells were infected with SPBN, SPBN-Gag, SPBN-333-Gag, SPBN-ΔCD-Gag, SPBN-333-ΔCD-Gag, BNSP, or BNSP-Gag (MOI of 2) and lysed 48 h later. Proteins were separated by SDS-PAGE and subjected to Western blotting with antibodies specific for RV RNP (A), RV G (B), or HIV-1 p24 (C). A protein of the expected size for RV N was detected for all recombinant RVs (A), whereas an RV G-specific antibody detected a 62-kDa protein for SPBN, SPBN-Gag, SPBN-333-Gag, BNSP, and BNSP-Gag and a slightly faster migrating protein in cell extracts from SPBN-ΔCD-Gag- and SPBN-333-ΔCD-Gag-infected cells (B). Expression of HIV-1 p55 from SPBN-Gag, SPBN-333-Gag, SPBN-ΔCD-Gag, SPBN-333-ΔCD-Gag, and BNSP-Gag was confirmed with an HIV-1 Gag-specific antibody.

FIG. 3.

Multicycle replication and one-step growth curve of recombinant vaccine vectors. BSR cells were infected with SPBN, SPBN-Gag, SPBN-333-Gag, SPBN-ΔCD-Gag, SPBN-333-ΔCD-5, BNSP, or BNSP-Gag at a MOI of 0.01 (multicycle growth; A) or 5 (one-step growth curve; B). Aliquots of tissue culture supernatants were collected, and viral titers were determined in duplicate.

FIG. 4.

New RV-based vaccine vectors are safe even after i.c. inoculation in mice. Six- to 8-week-old Swiss-Webster mice were inoculated i.c. with 10⁵ infectious particles of SPBN, SPBN-Gag, SPBN-333-Gag, SPBN-ΔCD-Gag, SPBN-333-ΔCD-Gag, BNSP, or BNSP-Gag. Mice were observed for 4 weeks, and average weights (A) and mortalities (B) were recorded. Error bars (A), standard deviations.

FIG. 5.

Surviving mice develop a strong immune response against RV G. Surviving mice were bled 28 days after i.c. injection, and pooled sera from each group were analyzed for ELISA reactivity against RV G in serial dilutions. OD, optical density.

FIG. 6.

Recombinant vaccine vehicles induce similar humoral and cellular immune responses in immunized mice. (A) RV G ELISA. Groups of five BALB/c mice were immunized with SPBN, SPBN-Gag, SPBN-333-Gag, SPBN-ΔCD-Gag, SPBN-333-ΔCD-Gag, BNSP,BNSP-Gag, or VSV-Gag and bled 3 weeks after immunization. Serial dilutions from pooled sera from each immunization group were used to determine the antibody titers against RV G by ELISA. 1 to 9 on the x axis indicates twofold dilutions from 1:100 to 1:25,600. (B) ELISPOT assays. Groups of two BALB/c mice were immunized with SPBN, SPBN-Gag, SPBN-333-Gag, SPBN-ΔCD-Gag, SPBN-333-ΔCD-Gag, BNSP, and BNSP-Gag and challenged 4 to 5 weeks after immunization with a recombinant vaccinia virus expressing HIV-1 Gag. Pooled splenocytes from two mice were analyzed at different dilutions for cells secreting IFN-γ. Error bars, standard deviations from three independent experiments. The recombinant VSV expressing HIV-1 Gag (VSV-Gag) was included in the last round of experiments (no error bar). OD, optical density.